Anthropic Acquires Coefficient Bio in $400M Deal

Anthropic has acquired stealth biotech AI startup Coefficient Bio in an all‑stock transaction valued at roughly $400 million, bringing a team of fewer than 10 researchers into its health and life‑sciences effort. Coefficient Bio, founded about eight months ago by former Genentech computational biology researchers Samuel Stanton and Nathan C. Frey, worked on AI tools for drug discovery and R&D planning while remaining in stealth with no disclosed product or revenue. Investors including Dimension held a large stake; one report cited an outsized internal rate of return for that investor on the sale. The deal underscores Anthropic’s push into life sciences and the growing trend of AI companies acquiring specialized biotech teams.
What happened
Anthropic has purchased Coefficient Bio, a stealth biotech AI startup, in an all‑stock transaction valued at just over $400 million. The acquisition was reported April 3–5, 2026 and brings a compact team of roughly nine to ten employees into Anthropic’s health and life‑sciences organization.
Context: Coefficient Bio was founded about eight months before the sale by Samuel Stanton and Nathan C. Frey, both previously affiliated with Genentech’s Prescient Design computational drug‑discovery unit. The company remained in stealth, with no publicly disclosed product or revenue, and positioned itself to apply generative and other machine‑learning approaches to accelerate drug discovery, draft R&D plans, and manage clinical and regulatory strategies. Industry coverage frames the transaction as part of a broader wave of leading AI companies expanding into biotech and life sciences research.
Key details
Multiple reports characterize the deal as an all‑stock transaction valued at approximately $400 million. Coefficient Bio’s team—described in reports as fewer than 10 staffers—will join Anthropic’s life‑sciences group. Reporting also notes that Dimension, a New York‑based venture firm, held roughly half of Coefficient Bio and realized a very large internal rate of return on the exit; one outlet cited a 38,513% IRR for that investor. Coverage highlights the founders’ academic and publication credentials, and the company’s orientation toward foundation‑model approaches to biological problems. Observers emphasize that Anthropic had already signaled interest in scientific research tools in prior product activity and is now deploying capital to build in‑house biological research capability.
Implications: The acquisition signals that Anthropic is committing material resources to integrate specialized computational biology talent and capabilities into an AI company that is primarily known for general‑purpose models. For the industry, the deal exemplifies how AI platform developers are obtaining expertise and research teams through acquisitions rather than solely through organic hiring or partnerships, accelerating the convergence of large‑model capabilities and domain‑specific biological research. For investors, the deal has already produced headline‑grabbing returns for early backers, prompting renewed attention on the valuation dynamics for tiny, research‑heavy startups in the AI‑biotech intersection.
What to watch next
Monitor how Anthropic integrates Coefficient Bio personnel into product roadmaps, whether the company discloses an explicit life‑sciences research product or platform, and whether regulators or customers raise questions about AI‑driven discovery workflows. Also watch for competitive responses from other large AI firms and for announcements from Dimension or the founders about future research outputs or organizational roles.
Scoring Rationale
Multiple reputable outlets corroborate an all‑stock acquisition valued at about $400M and a team of fewer than 10, boosting credibility and relevance to AI/biotech. The story scores high for novelty and sector relevance but loses one point for being a very recent (1–3 day old) transaction and limited immediate actionable detail.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
